Compressed Sensing-Based Channel Estimation for Intelligent Reflective Surface Aided mmWave Communication

Intelligent Reflective surface(IRS) aided communication is regarded as one of the key enabling technologies to improve the performance of future wireless networks. However, its channel state information(CSI) acquisition remains a significant challenge due to its large number of passive elements that lack signal processing and decoding capabilities. As a result, estimation of such large parameters of the cascaded channel from UE to the surface and from the surface to base station becomes more difficult than in conventional systems. In this paper, we begin by describing intelligent reflective surface assisted mmWave multiple-input single-outpu(MISO) channel model. Sebsequently, we propose Multiple Measurement based Compressive Sensing(CS) channel estimation approach that exploits structural sparsity induced bymmWave channel behavior. Further, we consider signal subspace of the observation data which is commonly attributed to Multiple Signal Classification(MUSIC) algorithm, to enhance the channel estimation accuracy by leveraging the diversity gain inherent in the structurally sparse channels. Our proposed approach showed a lower complexity compared to other existing estimation techniques used in the literature. Simulation results demonstrate its improved normalized mean square error (NMSE) performance relative to both Single Measurement Vector (SMV) and other MMV-based estimation methods.